Wearable Wireless Sensor for Multi-Scale Physiological Monitoring

Abstract

One of the aims of Year 1 of the project was to develop a prototype multi-channel pulse oximeter that can be used to collect physiological data from multiple body locations to combat motion artifact contamination. Specifically, the aim was to investigate if a motion artifact-free signal can be obtained in at least one of the multi-channels at any given time. Towards this aim, we have developed a prototype 6-photodetector reflectance-based pulse oximeter and preliminary results show that good signals can be obtained in one of the multi-channels at any given time. A conference proceedings paper describing detailed results is provided with the annual report. The second major aim of the project was to develop a motion and noise detection algorithm and a separate algorithm for the reconstruction of the motion and noise contaminated portion of the data. For detection of motion and noise artifacts, we have successfully developed an accurate and real-time realizable algorithm. Moreover, our detection algorithm is able to discriminate between severely and moderately corrupted data. The sensitivity and specificity of detecting severely corrupted data was found to be 98.7% and 92.9%, respectively. The sensitivity and specificity of detecting moderately corrupted data was found to be 94.4% and 90.4%, respectively. Comparison of our detection algorithm to some of the gold standard algorithms showed that our algorithm is far superior to the latter methods. For reconstruction of the motion and noise corrupted data segments, we have successfully developed an algorithm which significantly outperforms a gold standard method. It was found that our reconstruction algorithm consistently provides accurate estimations of the reconstructed motion and noise artifact contaminated signal s heart rates and oxygen saturation values when verified with the reference signal.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 2013
Accession Number
ADA590832

Entities

People

  • Ki Chon
  • Yitzhak Mendelson

Organizations

  • Worcester Polytechnic Institute

Tags

Communities of Interest

  • Biomedical
  • Engineered Resilient Systems
  • Human Systems
  • Sensors

DTIC Thesaurus Topics

  • Algorithms
  • Computational Science
  • Computer Programming
  • Detection
  • Detectors
  • Health Services
  • Heart Rate
  • Information Science
  • Machine Learning
  • Measurement
  • Medical Personnel
  • Oxygenation
  • Physiological Monitoring
  • Reliability
  • Smartphones
  • Statistical Analysis
  • Supervised Machine Learning

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Image Processing and Computer Vision.
  • Oncology and Biomarker-Based Cancer Detection.